Open Peer Review. Open Publishing. Open Access. Open Discussion. Open Directory. Open Recommendations. Open API. Open Source.
Universal Successor Representations for Transfer Reinforcement Learning
Chen Ma, Junfeng Wen, Yoshua Bengio
Feb 12, 2018 (modified: Feb 12, 2018)ICLR 2018 Workshop Submissionreaders: everyone
Abstract:The objective of transfer reinforcement learning is to generalize from a set of previous tasks to unseen new tasks. In this work, we focus on the transfer scenario where the dynamics among tasks are the same, but their goals differ. Although general value function (Sutton et al., 2011) has been shown to be useful for knowledge transfer, learning a universal value function can be challenging in practice. To attack this, we propose (1) to use universal successor representations (USR) to represent the transferable knowledge and (2) a USR approximator (USRA) that can be trained by interacting with the environment. Our experiments show that USR can be effectively applied to new tasks, and the agent initialized by the trained USRA can achieve the goal considerably faster than random initialization.
Keywords:Reinforcement Learning, Transfer Learning, Deep Learning, Successor Representations
Enter your feedback below and we'll get back to you as soon as possible.